Multi-Task Learning with Covariate-Overlap Regularization

Yang Sui, Qi Xu, Yang Bai +1

Multi-task learning improves data efficiency by sharing information across related tasks, but indiscriminate sharing can be harmful when their covariate distributions and response…